Machine Learning Classification Model for Identifying Pornography Addiction Among Children

Abstract

Through the development of civilization, there are different addictive behaviour. With the new technologies living in our daily life, Internet addiction has shown growth, especially for children. Among all the Internet addiction behaviours, porn addiction may raise attention to people since it may cause children with learning disabilities, depression, and social skills. Usually, the psychologist will identify the children with porn addiction with the results of the questionnaire and third-party observers. These methods depend on the first-party experience, and they may often overlook. From the review, we know that addictive behaviour is detectable by brain activity. Some research found that functional magnetic resonance imaging (fMRI) and electroencephalogram (EEG) are able to display the difference for addictive subjects. Hence, we want to use the EEG signals to identify pornography addiction among children. We use the bandpass as the preprocessing method. Wavelet Packet Decomposition (WPD) as feature extraction, and support vector machine (SVM) as classification methods to process the data to do the prediction model. This research is related to the government campaign under the MCMC, namely "Klik dengan bijak.".